Bug 12582 - Crash on GeForce 7300 SE
Summary: Crash on GeForce 7300 SE
Status: RESOLVED INVALID
Alias: None
Product: xorg
Classification: Unclassified
Component: Driver/nouveau (show other bugs)
Version: git
Hardware: x86-64 (AMD64) Linux (All)
: high major
Assignee: Nouveau Project
QA Contact: Xorg Project Team
URL:
Whiteboard:
Keywords: NEEDINFO
Depends on:
Blocks:
 
Reported: 2007-09-26 16:01 UTC by Andrey V. Panov
Modified: 2008-02-01 11:35 UTC (History)
1 user (show)

See Also:
i915 platform:
i915 features:


Attachments
output from lspci -vv (3.39 KB, application/x-bzip2)
2007-09-26 16:02 UTC, Andrey V. Panov
no flags Details
Xorg.0.log (23.60 KB, application/x-bzip2)
2007-09-26 16:03 UTC, Andrey V. Panov
no flags Details
xorg.conf (5.56 KB, application/x-bzip2)
2007-09-26 16:03 UTC, Andrey V. Panov
no flags Details
Xorg.0.log made without nvidiafb (13.83 KB, application/x-bzip2)
2007-09-27 16:02 UTC, Andrey V. Panov
no flags Details
screenshot of artifacts (132.83 KB, image/png)
2007-09-27 16:03 UTC, Andrey V. Panov
no flags Details
the screenshot of glxgears (128.33 KB, image/png)
2007-11-05 18:42 UTC, Andrey V. Panov
no flags Details

Description Andrey V. Panov 2007-09-26 16:01:36 UTC
After start X displays artifacts on a screen, mostly at its bottom. Switching to console causes crash and lock of a kernel with some responsibility to magic sysrq. I used yesterday git versions of drm, xf86-video-nouveau and mesa on slamd64-current linux (64 bit) distribution, linux 2.6.22.7 with SMP support on Athlon dual core CPU, nvidiafb module was loaded.
Comment 1 Andrey V. Panov 2007-09-26 16:02:37 UTC
Created attachment 11772 [details]
output from lspci -vv
Comment 2 Andrey V. Panov 2007-09-26 16:03:16 UTC
Created attachment 11773 [details]
Xorg.0.log
Comment 3 Andrey V. Panov 2007-09-26 16:03:47 UTC
Created attachment 11774 [details]
xorg.conf
Comment 4 Stephane Marchesin 2007-09-26 16:20:26 UTC
Could you please retry without the nvidiafb module (this is not supported) and provide a kernel log ? 
Comment 5 Andrey V. Panov 2007-09-27 16:00:51 UTC
Without nvidiafb linux kernel did not hang. By the way other drm modules (ati) can work simultaneously with framebuffer. 

After start there were artifacts on a screen, moving of cursor or windows caused changes in the artifacts. 2D speed was (for example moving a window in KDE) much lower than for "nv" driver. xdriinfo displayed nouveau driver, glxinfo showed direct rendering. Start of glxgears opened black window, screen was locked (no reaction on mouse or keyboard), then after several seconds X crashed.

After that /var/log/messages has following:

Sep 27 18:31:44 darkstar kernel: [drm] Allocating FIFO number 0
Sep 27 18:31:44 darkstar kernel: [drm] nouveau_fifo_alloc: initialised FIFO 0
Sep 27 18:31:44 darkstar kernel: [drm] Allocating FIFO number 1
Sep 27 18:31:44 darkstar kernel: [drm] nouveau_fifo_alloc: initialised FIFO 1
Sep 27 18:31:58 darkstar kernel: [drm] Allocating FIFO number 2
Sep 27 18:31:58 darkstar kernel: [drm] nouveau_fifo_alloc: initialised FIFO 2
Sep 27 18:31:58 darkstar kernel: [drm] nouveau_fifo_free: freeing fifo 2
Sep 27 18:32:22 darkstar kernel: [drm] Allocating FIFO number 2
Sep 27 18:32:22 darkstar kernel: [drm] nouveau_fifo_alloc: initialised FIFO 2
Sep 27 18:32:28 darkstar kernel: [drm] nouveau_fifo_free: freeing fifo 1
Sep 27 18:32:28 darkstar kernel: [drm] nouveau_fifo_free: freeing fifo 2
Sep 27 18:32:28 darkstar kernel: [drm] nouveau_fifo_free: freeing fifo 0

and /var/log/syslog:

Sep 27 18:32:22 darkstar kernel: [drm:nouveau_pgraph_irq_handler] *ERROR* PGRAPH error interrupt
Sep 27 18:32:22 darkstar kernel: [drm:nouveau_graph_dump_trap_info] *ERROR* nSource: DATA_ERROR, nStatus: BAD_ARGUMENT
Sep 27 18:32:22 darkstar kernel: [drm:nouveau_graph_dump_trap_info] *ERROR* Channel 2/0 (class 0x0062) - Method 0x0304, Data 0x00000000:0x15e004c0

Comment 6 Andrey V. Panov 2007-09-27 16:02:21 UTC
Created attachment 11790 [details]
Xorg.0.log made without nvidiafb
Comment 7 Andrey V. Panov 2007-09-27 16:03:12 UTC
Created attachment 11791 [details]
screenshot of artifacts
Comment 8 Stephane Marchesin 2007-10-29 18:00:28 UTC
Something is using opengl (as testified by the kernel's "Allocating FIFO number 2"). IF you do not install nouveua_dri.so, do you still have 2D corruption (which is what I'm interested in fixing ATM).
Comment 9 Andrey V. Panov 2007-11-05 18:34:59 UTC
I tried to remove nouveua_dri.so and there are still artefacts on the screen.
Comment 10 Andrey V. Panov 2007-11-05 18:42:43 UTC
Created attachment 12364 [details]
the screenshot of glxgears

This screenshot was made with nouveau (mesa, drm) drivers as of 2007.10.31. It does not crash or cause kernel lock anymore. But the output of glxgears is very distorted. Morever, the new 2d artefact introduced: sometimes fonts are too bright (like in the gkrellm window on the screenshot).
Comment 11 Maarten Maathuis 2008-01-17 06:08:21 UTC
Were you the person with Graphic Context Size issue that came to the irc channel?
Comment 12 Maarten Maathuis 2008-02-01 11:35:16 UTC
Closing due to lack of information.


Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.